Received: 22 February 2018 | Revised: 21 June 2018 | Accepted: 23 July 2018

DOI: 10.1002/bit.26806

AR T I C L E

Contribution of cellulose synthesis, formation of fibrils andtheir entanglement to the self‐flocculation of Zymomonasmobilis

Juan Xia1 | Chen‐Guang Liu1 | Xin‐Qing Zhao1 | Yi Xiao2 | Xiao‐Xia Xia2 |Feng‐Wu Bai3

1State Key Laboratory of Microbial

Metabolism, Department of Bioengineering,

Shanghai Jiao Tong University, Shanghai,

China

2Joint International Research Laboratory of

Metabolic and Developmental Sciences,

Shanghai Jiao Tong University, Shanghai,

China

3School of Life Science and Biotechnology,

Shanghai Jiao Tong University, Shanghai,

China

Correspondence

Xin‐Qing Zhao, State Key Laboratory of

Microbial Metabolism, Department of

Bioengineering, Shanghai Jiao Tong

University, 800 Dongchuan Rd., Shanghai,

200240, China.

Email: xqzhao@sjtu.edu.cn

Feng‐Wu Bai, School of Life Science and

Biotechnology, Shanghai Jiao Tong University,

800 Dongchuan Rd., Shanghai, 200240, China.

Email: fwbai@sjtu.edu.cn

Funding information

National Natural Science Foundation of China,

Grant/Award Number: 21536006

Abstract

Due to the unique Entner‐Doudoroff pathway, Zymomonas mobilis has been acknowl-

edged as a potential host to be engineered for biorefinery to produce biofuels and

biobased chemicals. The self‐flocculation of Z. mobilis can make the bacterial cells self‐immobilized within bioreactors for high density to improve product productivities, and in

the meantime enhance their tolerance to stresses, particularly product inhibition and the

toxicity of byproducts released during the pretreatment of lignocellulosic biomass. In this

work, we explored mechanism underlying such a phenotype with the self‐flocculatingstrain ZM401 developed from the regular non‐flocculating strain ZM4. Cellulase de‐flocculation and the restoration of the self‐flocculating phenotype for the de‐flocculatedbacterial cells subjected to culture confirmed the essential role of cellulose biosynthesis

in the self‐flocculation of ZM401. Furthermore, the deactivation of both Type I and Type

IV restriction‐modification systems was performed for ZM4 and ZM401 to improve their

transformation efficiencies. Comparative genome analysis detected the deletion of a

thymine from ZMO1082 in ZM401, leading to a frame‐shift mutation for the putative

gene to be integrated into the neighboring downstream gene ZMO1083 encoding the

catalytic subunit A of cellulose synthase, and consequently created a new gene to encode

a larger transmembrane protein BcsA_401 for more efficient synthesis of cellulose as

well as the development of cellulose fibrils and their entanglement for the self-

flocculation of the mutant. These speculations were confirmed by the morphological

observation of the bacterial cells under scanning electron microscopy, the impact of the

gene deletion on the self‐flocculation of ZM401, and the restoration of the self‐flocculating phenotype of ZM401ΔbcsA by the gene complementation. The progress will

lay a foundation not only for fundamental research in deciphering molecular mechanisms

underlying the self‐flocculation of Z. mobilis and stress tolerance associated with the

morphological change but also for technological innovations in engineering non‐flocculating Z. mobilis and other bacterial species with the self‐flocculating phenotype.

K E YWORD S

cellulose synthesis, fibrils formation and entanglement, frame‐shift mutation, self‐flocculation,Zymomonas mobilis

Biotechnology and Bioengineering. 2018;115:2714–2725.wileyonlinelibrary.com/journal/bit2714 | © 2018 Wiley Periodicals, Inc.

1 | INTRODUCTION

Compared to the Embden‐Meyerhof‐Parnas pathway in Saccharomyces

cerevisiae with 2 mole ATP produced from 1 mole glucose metabolized,

Zymomonas mobilis ferments glucose through the Entner‐Doudoroffpathway with only 1 mole ATP produced, and consequently less biomass

is accumulated for high ethanol yield (Bai, Anderson, &Moo‐Young, 2008;Kalnenieks, 2006). But unfortunately, this species has never been used

for ethanol production in the industry due to its narrow substrate

spectrum. Ethanol production from lignocellulosic biomass provides

opportunities for exploring the advantage of Z. mibilis, because only

glucose is released from cellulose hydrolysis, and in the meantime this

species can be engineered with xylose isomerase to convert xylose to

xylulose directly without cofactor imbalance and xylitol accumulation that

are intrinsic to S. cerevisiae engineered with xylose reductase and xylitol

dehydrogenase for ethanol production from xylose (Jeffries, 2006; M.

Zhang, Eddy, Deanda, Finkelstein, & Picataggio, 1995). In addition, Z.

mobilis is also a potential host to be engineered for biorefinery to produce

biobased chemicals such as 2, 3‐butanediol (M. He et al., 2014; Yang, Fei

et al., 2016; Yang, Mohagheghi et al. 2016).

The regular morphology of Z. mobilis is characterized by single cells

measured at 2–3×1–2 μm, but the bacterial cells can self‐flocculate to

form flocs at a magnitude of hundreds of micrometers (Zhao et al., 2014).

From the viewpoint of engineering, the self‐flocculation of microbial cells

presents advantages if the self‐flocculating process is controlled properly

without significant mass transfer limitation. On the one hand, cell flocs

can be self‐immobilized within bioreactors for high density to improve

productivities of the bioreactors, particularly under continuous culture

and fermentation conditions, which was highlighted in ethanol production

by the self‐flocculating S. cerevisiae (Zhao & Bai, 2009) and wastewater

treatment in upflow anaerobic sludge blanket bioreactors using activated

sludge granules developed by the self‐flocculation of different micro-

organisms (Chong, Sen, Kayaalp, & Ang, 2012; Lim & Kim, 2014). On the

other hand, enhanced tolerance to stresses such as ethanol produced

during fermentation and acetic acid released from the hydrolysis of

hemicelluloses during the pretreatment of lignocellulosic biomass was

observed with the self‐flocculating strains from S. cerevisiae and Z. moblis

(Xue, Zhao, & Bai, 2010; Zhao et al., 2014). We hypothesize that

molecular mechanism underlying this phenomenon would be enhanced

quorum sensing (QS) associated with the morphological change,

particularly for Z. mobilis, because the self‐flocculation of microbial cells

characterized by cell‐to‐cell contact is the upper limit for high‐densityculture that is required for triggering QS (Papenfort & Bassler, 2016).

Although mechanism underlying the self‐flocculation of S. cerevisiae

has been elucidated for engineering other yeast strains with the

phenotype (L. Y. He, Zhao, & Bai, 2012; Q. Li, Zhao, Chang, Zhang, &

Bai, 2012), little has been done for the self-flocculation of Z. mobilis. S.

cerevisiae is eukaryotic, and extracellular polymeric substances (EPSs) for

its self‐flocculation are glycoproteins (Halme, Bumgarner, Styles, & Fink,

2004; Soares, 2010; Verstrepen, Jansen, Lewitter, & Fink, 2005; Zhao &

Bai, 2009), but Z. mobilis is prokaryotic, which lacks glycosylation function

for proteins. Although lot of work has been done with EPSs in biofilms

developed by pathogens and activated sludge for sewage treatment

(Flemming & Wingender, 2010; Liu et al., 2010), significant difference in

morphologies among biofilms, activated sludge and the bacterial flocs

indicates that EPSs flocculating Z. mobiliswould be different, or these EPS

components might interact through different mechanism. Moreover, in

addition, most biofilms and all activated sludge are formed by mixed

organisms through mutualistic symbiosis among different species

(Flemming et al., 2016; Ju & Zhang, 2015), which are also different from

the self‐flocculation of Z. mobilis under pure culture conditions. In

addition, Z. mobilis has evolved with restriction‐modification (R‐M)

systems such as Type I and Type IV, which not only protect it from the

invasion of foreign DNA but also impede genetic manipulations on the

species. Although deactivation of the single R‐M system was attempted

(Kerr, Jeon, Svenson, Rogers, & Neilan, 2011), the effort is still needed for

developing deactivated mutants to transform Z. mobilis more efficiently

for fundamental research and technological innovations.

In this work, the contribution of cellulose biosynthesis to the self‐flocculation of Z. mobilis was studied, and mutants with the dual R‐Msystems deactivated were developed. The progress will lay a

foundation for further fundamental research in deciphering mechan-

ism underlying the self‐flocculation of Z. mobilis and stress tolerance

associated with the morphological change as well as technological

innovations for engineering non‐flocculating Z. mobilis and other

bacterial strains with desired self‐flocculating phenotypes.

2 | MATERIALS AND METHODS

2.1 | Strains, media, and enzymes

Escherichia coli DH5α and E. coli JM110 was used for the propagation of

plasmids carrying genes with and without methylation, respectively, to

check the impact of the deactivation of the R-M system(s) on the

transformation of Z. mobilis, which were cultured with the Luria‐Bertanimedium composed of 10 g/L tryptone, 5 g/L yeast extract and 10 g/L

NaCl in flasks at 37°C and 200 rpm. Z. mobilis strains, the non‐flocculatingZM4, and self‐flocculating ZM401, were purchased from American Type

Culture Collection (ATCC), which were cultured in a rich medium

composed of 10 g/L yeast extract, 20 g/L glucose, and 2 g/L KH2PO4 at

30°C and 150 rpm. When 20 μg/ml tetracycline or 5% sucrose was

supplemented, the rich medium was used for positive‐ and counter‐section (Hmelo et al., 2015). All strains used in this study are given in

Table 1.

Enzymatic hydrolysis was designed to identify polysaccharides in

the EPSs for the self‐flocculation of ZM401, and enzymes that specially

hydrolyze different glycosidic bonds were selected, including α‐amylase

and glucoamylase donated by COFCO Nutrition and Health Research

Institute, China to hydrolyze α‐glycosidic bonds as well as cellulases

(Novozymes), endo‐β‐glucanase (Sigma), cellobiase and β‐glucosidase(Novozymes) to hydrolyze β–(1→4) linked glycosidic bonds.

2.2 | Calcofluor‐white staining assay

Calcofluor‐white staining was used to qualitatively evaluate cellulose

synthesized by Z. mobilis (Cowles, Willis, Engel, Jones, & Barak, 2015).

XIA ET AL. | 2715

The fluorescent dye was supplemented at 20 μg/ml into 1ml culture

with OD600 adjusted to 1.0, and the mixture was incubated at room

temperature under dark conditions for 1 hr, which was then

centrifuged at 12,000g for 1min to collect cell pellets, washed twice

with deionized (DI) water and re‐suspended to 1ml by DI water for

fluorescence analysis. The fluorescence density was measured by

ENSPIRE 2300 (PerkinElmer, Waltham, MA) at an excitation of

360 nm and an emission of 460 nm.

2.3 | Electron microscopy characterization

The morphology of Z. mobilis was observed under scanning electron

microscopy (SEM). Samples were prepared as previously reported with

minor modifications (Gu, Zhang, & Bao, 2015). Briefly, cells were

collected, washed with phosphate buffer solution (PBS), and then fixed

with 2.5% glutaraldehyde for 4 hr. After washing three times with PBS,

the cells were sequentially dehydrated by 50%, 70%, 90%, and 100%

ethanol for 15min, which was then subjected to critical point drying and

coated with gold for SEM observation using Hitachi S‐3400N.

2.4 | Characterization of the self‐flocculationof Z. mobilis

The flocculating efficiency of Z. mobilis was quantified by the method

reported previously with modifications on the de‐flocculation of the

bacterial flocs (Palha, Lopes, & Pereira, 1997). Simply, test tubes, each

with 5ml culture, were shaken vigorously for homogenous suspension,

and then rested for 5min. Suspension of 200 μl was sampled from the

upper section to measure OD600 to quantify the non‐flocculating cells

(Cf). Based on the previous study on the de‐flocculation of ZM401 by

cellulase, cellulase (Novozymes) at designated dosages was supplemen-

ted into the culture, which was incubated at 50°C for complete de‐flocculation of the bacterial flocs to measure the concentration of total

cells (Ct). The flocculating efficiency (R%) of Z. mobilis was evaluated by

the calculation: R%= (1 −Cf/Ct) × 100.

2.5 | Genetic manipulations

Plasmids and primers used in this study are given in Table 2 and Table 3.

Gene deletion was performed by homologous recombination with the

suicide vector pEX18Tc bearing tetracycline selection marker (tetA) and

sucrose counter‐selection marker (sacB, which are illustrated schema-

tically in Figure 1a and b, respectively. Briefly, 500–1,000 bp fragments

flanking gene to be deleted were amplified, and fused by overlap

extension polymerase chain reactions (SOE‐PCR). The final PCR

products were purified, and digested with restriction enzymes for

cloning into the suicide plasmid to develop recombinant plasmids, which

TABLE 1 Strains used in this study

Strains Description References

E. coli DH5α lacZΔM15, recA1

E. coli JM110 rpsL, dam‐, dcm‐

Z. mobilis ZM4 ATCC31821

Z. mobilis

ZM401

ATCC31822

ZM401ΔbcsA ZM401 with bcsA deleted This study

ZM401ΔbcsA/bcsA_ZM4

ZM401ΔbcsA complemented with

bcsA from ZM4

This study

ZM401ΔbcsA/bcsA_401

ZM401ΔbcsA complemented with

bcsA_401 from ZM401

This study

ZM401ΔbcsA/bcsA_401R575A

ZM401ΔbcsA complemented with

bcsA_401 R575A

This study

ZM4m ZM4 deleted with ZMO0028 and

ZMO1933

This study

ZM401m ZM401 deleted with ZMO0028

and ZMO1933

This study

TABLE 2 Plasmids used in this study

Plasmids Description References

pEX18Tc TcR, oriT+, sacB+, gene replacement vector with

MCS from pUC18

Hoang et al. (1998)

pHW20a TcR, mob (RP4), mob (RSF1010), lacZα, MCS

and oriV

Dong et al. (2011)

pEX18Tc::Δ1933 Fragment flanking ZMO1933 of ZM4 ligated

into pEX18Tc

This study

pEX18Tc::Δ0028 Fragment flanking ZMO0028 of ZM4 ligated

into pEX18Tc

This study

pEX18Tc::ΔbcsA Fragment flanking bcsA of ZM401 ligated into

pEX18Tc

This study

pHW20a:: bcsA bcsA from ZM4 ligated into pHW20a with the

pdcZm promoter

This study

pHW20a::bcsA_401 bcsA_401 from ZM401 ligated into pHW20a

with the pdcZm promoter

This study

pHW20a::

bcsA_401R575AMutated bcsA_401 from ZM401 ligated into

pHW20a with the pdcZm promoter

This study

2716 | XIA ET AL.

were propagated in E. coli DH5α, and confirmed by sequencing. Due to

the R‐M systems in Z. mobilis, the confirmed plasmids were further

transformed into E. coli JM110 for demethylation and more efficient

transformation into ZM4 and ZM401 through electroporation by the

Bio‐Rad Gene Pulser with 0.1 cm gap cuvettes operated at 16 kV/cm,

200Ω and 25 μF. The positive selection was performed using the rich

medium supplemented with tetracycline after the first crossover

recombination. Then, the second crossover recombination was per-

formed, and a mutant with the gene deleted was selected through the

counter‐selection using the rich medium supplemented with sucrose (X.

Li, Thomason, Sawitzke, Costantino, & Court, 2013), which was

confirmed by PCR. For gene overexpression/complementation, promo-

ters and target genes were amplified separately by PCR. The

recombinant plasmids were constructed through inserting promoters

and target genes into the shuttle vector pHW20a (Figure 1c) by

seamless cloning (Dong, Bao, Ryu, & Zhong, 2011), which were

propagated in E. coli DH5α and E. coli JM110 and confirmed by

sequencing. The transformation procedures of these plasmids were the

same as that applied with the plasmids used for the gene deletion.

2.6 | RNA extraction and transcription analysis

Cells were harvested at the exponential growth stage. Total RNA was

extracted using the Spin Column Bacterial total RNA Purification Kit

(Sango, Shanghai, China) according to the manufacturerʼs instructions.

One microgram RNA was used as a template to synthesize cDNA using

the PrimeScript™ RT reagent cDNA Kit (Takara, Japan). Gene

transcription was quantified by quantitative PCR (qPCR) with the

CFX System (Biorad, Hercules, CA), and the transcription of 16s RNA

was used as the reference for normalization (K. Zhang et al., 2015).

2.7 | Evaluation of cell growth and ethanolfermentation

Stock culture was transferred from agar plates using an inoculating

loop, and inoculated into 250ml Erlenmeyer flasks containing 100ml

rich medium for seed culture, which was performed overnight in a

shaker at 30°C and 150 rpm to exponential phase for inoculation at

20% (v/v) into 250ml Erlenmeyer flasks containing 100ml medium

composed of 100 g/L glucose, 10 g/L yeast extract, and 2 g/L

KH2PO4. The culture and ethanol fermentation was also performed

in shakers at 30°C and 150 rpm, and time‐courses were monitored by

sampling at designated intervals.

Cell growth was characterized by OD600. For the self‐flocculatingZ. mobilis strains, the de‐flocculation of the bacterial flocs by cellulase

was performed before the measurement of OD600. After the

measurement of OD600, the samples were centrifuged at 12,000g

for 1min to remove cells, and the supernatant was filtered by

0.22 μm membrane for the analysis of ethanol and residual glucose

using HPLC (Waters e2695 with a Bio‐Rad Aminex HPX‐87H organic

TABLE 3 Primers used in this study

Primers Sequences Description

Δ0028‐up‐F CGGAATTCGGCGGTATTTTGCTCCAGTC Deletion of ZMO0028 in ZM4 or

ZM401Δ0028‐up‐R GAAATCCAAATCAACCCGTTTGCTGGGAACCGCCATTATCG

Δ0028‐down‐F CGATAATGGCGGTTCCCAGCAAACGGGTTGATTTGGATTTC

Δ0028‐down‐R CGGGATCCAACGGCTGACCAAAACTATCC

Δ1933‐up‐F ACGGGATCCTCGGCTTCGACGTCCTTATG Deletion of ZMO1933 in ZM4 or

ZM401Δ1933‐up‐R GGTCAAGGTAAGCCATGGCGCAGATCGCTTAGGCGATCTTG

Δ1933‐down‐F CAAGATCGCCTAAGCGATCTGCGCCATGGCTTACCTTGACC

Δ1933‐down‐R TAACTGCAGGGTTGCAGCCGTCGTAGAAC

Δ1083‐up‐F CGGAATTCGTCTGTCGAAACGATGGATG Deletion of bcsA_401 in ZM401

Δ1083‐up‐R GATGCGATGGTTGAAACAGGATAGAGGCAAAATGGCTTGG

Δ1083‐down‐F CCAAGCCATTTTGCCTCTATCCTGTTTCAACCATCGCATC

Δ1083‐down‐R ACGGGATCCCAAATCTGTCAAAGAAAGGC

Ppdc‐F CGGCCAGTGAATTCGAGCTCTTATCGCTCATGATCGCGGC Amplification of the pyruvate

decarboxylase promotorPpdc‐R TGCTTACTCCATATATTCAAAAC

1082‐F GTTTTGAATATATGGAGTAAGCAATGCTTCATAAAAGCCGTAT Amplification of bcsA_401

1083‐F CTTAATAAGTTAGGAGAATAAACATGGGATTGACGCTGCTGAT Amplification of ZMO1083

1083‐R ATTACGCCAAGCTTGCATGCTCATTTCTTTACTGCCCCCT

bcsA_401R575A‐up‐R GGCGATCTGCCTACTTTCATGGGCGA Substitution of the 575th

Arginine of bcsA_401 by AlaninebcsA_401R575A‐down‐F TCGCCCATGAAAGTAGGCAGATCGCCAATAC

Note. The underlines indicated homologous sequences for fusion into adjacent fragments, and the black bold character indicated restriction enzymes sites.

XIA ET AL. | 2717

acids column and refractive index detector 2414). The oven

temperature was set at 65°C and 4mM H2SO4 (0.6 ml/min) was

used as the mobile phase.

3 | RESULTS AND DISCUSSION

3.1 | Identification of the role of cellulose in theself‐flocculation of ZM401

Previous studies assumed that cellulose could be the major EPS for

the self‐flocculation of ZM401 (Jeon, Xun, Su, & Rogers, 2012).

Moreover, protease treatment could not de‐flocculate the bacterial

flocs, which not only ruled out the possibility for proteins including

flagella to flocculate the bacterial cells but also indicated the role of

polysaccharides in the self‐flocculation of the bacterial cells. Initially,

we tried to characterize the cellulose component directly in the EPS

extracted from the self‐flocculated bacterial cells, but unfortunately,

we were not successful due to the low EPS content for extraction.

Therefore, we had to look for alternatives and turned to enzymatic

hydrolysis of the EPS for indirect evidence to confirm the role of

cellulose in the self‐flocculation of ZM401.

The impact on the self‐flocculation of ZM401 when treated by

different carbohydrate hydrolytic enzymes is shown in Figure 2a.

Both α‐amylase and glucoamylase could not de‐flocculate the flocs,

which ruled out the possibility for the bacterial cells to be flocculated

through polysaccharides linked by α‐glycosidic bonds either starch or

dextrin, particularly dextrin that is glutinuous. Endo‐β‐glucanase that

specifically hydrolyzes β–(1→4) linked glycosidic bonds into cello-

dextrin deflocculated the bacterial flocs significantly but not

completely as that observed in the treatment by cellulases. Both

cellobiase and β‐glucosidase alone could not de‐flocculate the

bacterial flocs, but when the two enzymes were supplemented

sequentially into the sample treated by endo‐β‐glucanase, the

complete de‐flocculation of ZM401 occurred. These experiments

supported our conclusion that cellulose rather than other poly-

saccharides flocculates the bacterial cells.

Figure 2b further quantifies the de‐flocculation of ZM401 by

cellulases supplemented at the dosages of 0.30 U/ml and 0.03 U/ml,

respectively, into the suspension adjusted to the same cell density of

OD600 = 2.0 detected after the flocs were de‐flocculated completely.

As can be seen, the bacterial flocs were de‐flocculated effectively by

cellulase, particularly when cellulase was supplemented at the higher

dosage of 0.30 U/ml, and a complete de‐flocculation was observed

within 10min. Under the lower cellulase dosage of 0.03 U/ml, the

incubation time was extended to 30min for a complete de‐flocculation. From the viewpoint of the intrinsic kinetics of enzymatic

catalysis and mass transfer associated with the heterologous

reaction, the reasons for this phenomenon might be: (a) lower

F IGURE 1 Strategies for homologousrecombination (a) and development ofvectors for gene deletion (b) and

expression/complementation (c) [Colorfigure can be viewed atwileyonlinelibrary.com]

2718 | XIA ET AL.

maximal rate associated with the lower uploading of cellulase, and (b)

lower driving force in mass transfer for cellulase to diffuse into the

inside of the bacterial flocs.

On the other hand, the de‐flocculated bacterial cells restored their

self‐flocculating phenotype when cellulase was removed by decanting

the supernatant after the suspension was centrifuged, and the cell

pellets were re‐suspended in the fresh medium and cultured at 30°C

for cellulose regeneration, because flocculating efficiency as high as

90% was detected after culture for 90min, but the increase in cell

numbers characterized by OD600 was only 43.5% (Figure 3a), indicating

that contribution to the self‐flocculating phenotype was mainly from

cellulose regeneration by the de‐flocculated bacterial cells rather than

their propagation. Such speculation was further confirmed by the

morphological difference observed in the SEM images, because smooth

surfaces were observed for the bacterial cells subjected to cellulase

treatment, but cellulose fibrils were observed on the surfaces of

ZM401 and the de‐flocculated bacterial cells subjected to the culture

(Figure 3b−d). Although cellulose has been reported as one of the

architectural elements of bacterial biofilms (Flemming et al., 2016;

Serra, Richter, & Hengge, 2013), its role in the self‐flocculation of

Z. mobilis is significantly different, because the entanglement of

cellulose fibrils was clearly observed in the SEM image. These results

not only confirmed the role of extracellular cellulose fibrils and their

entanglement in the self‐flocculation of ZM401 but also provided a

strategy for restoring the self‐flocculating phenotype once the bacterial

flocs are de‐flocculated by cellulase.

3.2 | Deactivation of the R‐M systems

Z. mobilis has evolved with R‐M systems for defense against invasion

of foreign DNA, which should be deactivated properly for efficient

genetic manipulations. However, physiological properties and meta-

bolic performance should not be affected substantially so that the R‐M system deactivated mutants are suitable for fundamental research

and technological innovations.

Transformation efficiencies of both methylated and unmethy-

lated plasmids were increased when ZMO1933 and ZMO0028

related to Type I and Type IV R‐M systems in Z. mobilis ZM4 were

knocked out separately (Kerr et al., 2011), but the researchers could

not obtain mutants with both I and IV R‐M systems deactivated due

to the lack of suitable selective markers. Herewith, we deactivated

both Type I and Type IV R‐M systems in ZM4 and ZM401 to further

improve their transformation efficiencies (Figure 4a,b), which were

examined by the transformation of the shuttle vectors pHW20a and

pHW20a (Dam‐, Dcm‐). The growth and ethanol fermentation

performance of the mutants ZM4m and ZM401m are shown in

Figure 4c,d. In addition, we compared the transformation efficiency

of ZM4 with both the R‐M systems deactivated and the deactivation

of the R‐M system either Type I or Type IV.

While almost no transformant was obtained when the wide‐typeZM4 and ZM401 were transformed by pHW20a amplified in E. coli

DH5α, the transformation efficiencies of 2.1 × 104/μg(DNA) and

1.2 × 104/μg(DNA) were achieved, respectively, for the mutants of ZM4

and ZM401 with both the R‐M systems deactivated. When the plasmid

was amplified in E. coli JM110, the transformation efficiencies were

8.0 × 102/μg(DNA) and 3.3 × 102/μg(DNA) for the wide‐type ZM4 and

ZM401, but were increased drastically to 2.4 × 104/μg(DNA) and

1.5 × 104/μg(DNA) after the R‐M systems were deactivated. Moreover,

higher transformation efficiency was observed for ZM4 with both the

R‐M systems deactivated compared to the deactivation of either of them.

No significant impact was observed in cell growth, glucose consumption

and ethanol production between the wide‐type strains and their mutants

with the R‐M systems deactivated. On the other hand, the deactivation

of the R‐M systems in ZM401 did not disrupt its self‐flocculationphenotype. Therefore, the mutants with both the R‐M systems

deactivated would benefit to experimental studies of molecular

mechanism underlying the self‐flocculation of ZM401 and other genetic

modifications on Z. mobilis.

3.3 | Mutations in genes responsible for cellulosesynthesis in ZM401

The comparative transcriptome analysis between ZM401 and ZM4

revealed differential expression of genes in ZM401: the expression of

F IGURE 2 Impact of enzymatic treatment on the

self‐flocculation of ZM401. (a) Incubated at 84°C withoutsupplementation of any enzyme as the control (1), and treatment of2 hr at 84°C by α‐amylase (2), 60°C by glucoamylase (3), and 50°C

by cellulases (4), endo‐β‐1, 4‐glucanse (5), cellobiase (6),β‐glucosidase (7), and endo‐β‐1,4‐glucanse+cellobiase+β‐glucosidase(8). All enzymes were supplemented with overdosages.

De‐flocculation of ZM401 by cellulase supplemented at 0.03 U/mland 0.30 U/ml, respectively (b). The initial cell density was adjustedto OD600 ≅2.0 which was measured after the bacterial flocs were

completely de‐flocculated. The enzymatic hydrolysis was performedat 50°C [Color figure can be viewed at wileyonlinelibrary.com]

XIA ET AL. | 2719

ZMO1083 and ZMO1084 that might relate to cellulose biosynthesis

was upregulated, but the expression of ZMO0608, ZMO0611,

ZMO0612, ZMO0613 and ZMO0614 for the synthesis of flagella‐related proteins was downregulated, indicating that enhanced

cellulose production and reduced flagella activities could facilitate

the self‐flocculation of Z. mobilis (Jeon et al., 2012). Cellulose is

different from other polysaccharides that are sticky to glue bacterial

cells together and also makes them adhered to surfaces of supporting

materials for biofilms formation (Limoli, Jones, & Wozniak, 2015),

which flocculated ZM401 through the entanglement of cellulose

fibrils (Figure 3b,d). Such a cellulose matrix is porous for more

efficient mass transfer, which was validated experimentally by our

previous studies on the growth and ethanol fermentation perfor-

mance of ZM401 (Zhao et al., 2014). On the other hand, the potential

reduction in the synthesis of flagellum proteins could not be the root

reason for the self‐flocculation of ZM401, because the function of

flagella is mainly for bacterial cells to swim rather than glue or

entangle together for their self‐flocculation.We sequenced the genome of ZM401 (Zhao, Bai, Zhao, Yang, &

Bai, 2012), and compared to that of ZM4 previously reported (Seo

et al., 2005; Yang et al., 2009). A deletion of nucleotide (Thymine) in

ZMO1082 was detected in ZM401, which consequently created a

frame‐shift mutation as illustrated in Figure 5a. In ZM4, ZMO1082 is

a putative gene, which is predicted to encode a peptide composed of

67 amino acid residues with unknown functions. It is less likely for

such a small peptide to be evolved into a protein with catalytic

functions. On the other hand, the bioinformatics analysis of

ZMO1083, ZMO1084 and ZMO1085 in ZM4 identified a putative

operon comprising of cellulose synthase catalytic subunits A, B and C

(BcsA, BcsB, and BcsC) for cellulose biosynthesis (Asraf, Rajnish, &

Gunasekaran, 2011), which show high homology to the cellulose

synthesis operon in Gluconacetobacter xylinus with the genome loci of

H845_RS02190, H845_RS02195 and H845_RS02200 for catalytic

subunits (Kubiak et al., 2014).

ZMO1083 encodes BcsA of cellulose synthases, which composes

of a UDP‐forming subunit and a C‐terminal PilZ domain, catalyzing

cellulose synthesis from UDP‐glucose upon binding to the second

messenger bis‐(3′,5′)‐cyclic guanosine monophosphate (c‐di‐GMP) for

signal transduction (Asraf et al., 2011; Jeon et al., 2012; Morgan,

McNamar, & Zimmer, 2015). The mutation of ZMO1082 in ZM401

disrupted the stop codon TGA for the translation of the putative

gene to be terminated as that in ZM4, but continued instead till

encountering the stop codon of the neighboring downstream gene

ZMO1083. With Open Reading Frame Finder (https://www.ncbi.nlm.

nih.gov/orffinder/), we predicted that the disrupted ZMO1082 in

ZM401 was integrated into ZMO1083 to create a new gene

bcsA_401 (Figure 5b), since an overlap of 23 nucleotides was

detected for the two genes. bcsA_401 is larger than ZMO1083, and

with TMHMM (http://www.cbs.dtu.dk/services/TMHMM/), we pre-

dicted that BcsA_401 encoded by this gene in ZM401 is a protein

bearing eight trans‐membrane helices, while the native BcsA in ZM4

was predicted with only six trans‐membrane helices (Figure 5c). We

compared this analysis with the sequences of amino acid residues

and membrane arrangements of the proteins in other cellulose‐producing bacteria reported in the NCBI database, and significant

homology was observed (date not shown).

Since cellulose synthases are cytoplasmic proteins and cellulose

synthesis is a membrane‐associated process (Mcnamara, Morgan, &

Zimmer, 2015; Morgan, Strumillo, & Zimmer, 2013), we predicted

F IGURE 3 Restoration of the self‐flocculating phenotype ofZM401 cells subjected to cellulase treatment (a) and morphologicaldifferences observed by SEM for ZM401 cells (b), ZM401 cells

subjected to cellulase treatment (c) and the culture of de‐flocculatedbacterial cells (d). SEM: scanning electron microscopy, and the redarrows indicate the entanglement of cellulose fibrils [Color figure can

be viewed at wileyonlinelibrary.com]

2720 | XIA ET AL.

that the difference in the trans‐membrane helices would affect their

catalytic activities for cellulose biosynthesis and export of synthe-

sized cellulose molecules. Based on the information of Z. mobilis in

the KEGG database (http://www.genome.jp/kegg/), ZMO1086 en-

codes a glycoside hydrolase for hydrolyzing β‐1,4‐glucosidic bond in

cellulose, which might work similar to another cellulose synthase

subunit BcsZ, and act synergistically with BcsA_ZM4/BcsA_401 to

regulate cellulose synthesis and degradation (Mazur & Zimmer,

2011). Although the mutation in ZMO1082 did not alter the reading

of their codons, the expression of bcsA_401 in ZM401 and altered

configurations and functions with BcsA_401 might affect the

transcription of its neighboring downstream genes bcsB and bcsC as

well as the functions of the proteins they encode. Such a speculation

was confirmed by the RT‐qPCR analysis, since the upregulation of

eight and six folds was observed respectively for the expression of

bcsB and bcsC in ZM401, compared to that detected in ZM4. In the

meantime, more significant difference in the transcription of

bcsA_401 (12‐fold upregulation) was also observed in ZM401

compared to that observed with the native bcsA in ZM4. The

enhanced transcription of the bcs operon in ZM401 would benefit its

cellulose synthesis, development of cellulose fibrils and their

entanglement for the self‐flocculating phenotype of ZM401, which

are illustrated schematically in Figure 5d.

The genome sequence comparison did not detect any mutations

in genes related to the synthesis of flagella proteins, which not only

supported our speculation on their role in the self‐flocculation of

ZM401, and but also indicates that in‐depth studies of molecular

mechanism underlying this phenotype should be focused on cellulose

biosynthesis, export, and development of cellulose fibrils as well as

how these processes are finely regulated for the self‐flocculation of

the bacterial cells.

3.4 | Role of bcsA_401 in the self‐flocculation ofZM401

To identify the role of bcsA_401 in the self‐flocculation of ZM401, the

knockout of ZMO1083, the major part of bcsA_401, was performed.

As expected, the mutant ZM401ΔbcsA lost the self‐flocculatingphenotype, and its flocculating efficiency was almost same as that

detected with ZM4, which was in accordance with the very weak

F IGURE 4 Impact of the deactivation of Type I and Type IV R‐M systems in ZM4 (a) and ZM401 (b) on their transformation efficiencies, andethanol fermentation by both the R-M systems de-activated mutants ZM4 (c) and ZM401 (d) [Color figure can be viewed at wileyonlinelibrary.com]

XIA ET AL. | 2721

F IGURE 5 Predicted genes related to cellulose synthesis in ZM4 and ZM401 (a), alignments of the amino acid sequences of the

proteins BcsA_ZM4 and BcsA_401 (b), their topological models analyzed by TMHMM and illustrated by TOPO2 (www.sacs.ucsf.edu/cgi‐bin/open‐topo2.py), in which the extra N‐terminal and essential sites RxxxR and NxSxxG for c‐di‐GMP binding are highlighted by orange, redand blue colors, respectively (c), and schematic diagram for the biosynthesis of cellulose fibrils in ZM401 (d). BcsA, BcsB, and BcsC: major

cellulose synthase operon proteins. GT: Glucosyltransferase; PilZ: c‐di‐GMP binding domain in the C‐terminal; UDP‐Glc: UDP‐glucose[Color figure can be viewed at wileyonlinelibrary.com]

2722 | XIA ET AL.

fluorescence intensity detected under the Calcofluor‐white staining

conditions (Figure 6a,b). SEM images showed that the cell surfaces of

ZM401ΔbcsA were smooth without cellulose fibrils (Figure 6c) as

that observed in ZM401 previously. These results indicated that the

deletion of bcsA_401 made the mutant lost the ability for synthesiz-

ing cellulose to flocculate the bacterial cells.

We further performed the gene complementation on

ZM401ΔbcsA under the direction of the native pyruvate decar-

boxylase promoter (Ppdc) from Z. mobilis (Conway, Sewell, &

Ingram, 1987). No flocculating transformants were selected when

bcsA_ZM4 was complemented into ZM401ΔbcsA. However,

when the whole bcsA_401 complementation was performed,

ZM401ΔbcsA restored the self‐flocculating phenotype (Figure 6a,

b). Morphological observation of the complemented bacterial cells

with SEM and cellulase treatment indicated and validated that the

self‐flocculation was due to the biosynthesis of cellulose, devel-

opment of cellulose fibrils and their entanglement (Figure 6c,d).

Analysis of BcsA_401 showed a conserved PilZ domain in the

C‐terminal characterized by RxxxR and NxSxxG (Figure 5c), which is

the binding site for c‐di‐GMP (Amikam & Galperin, 2006; Benach et al.,

2007). As a second messenger, c‐di‐GMP regulates multiple cellular

functions for development and morphogenesis, in particular the

synthesis of EPSs in bacteria for biofilm formation (Hengge, 2009;

Jenal, Reinders, & Lori, 2017). The PilZ domain of the G. xylinus

cellulose synthase BcsA binds with c‐di‐GMP, which acts as an

allosteric activator of the cellulose synthase, causing changes in the

configuration and conformation of the enzyme molecule, and conse-

quently initiates cellulose synthesis (Morgan et al., 2015). To explore

the role of the PilZ domain in BcsA_401, we constructed an expression

plasmid pHW20a::bcsA_401R575A with an alanine residue to substitute

the 575th arginine (AxxxR), and introduced the plasmid into

ZM401ΔbcsA for the gene complementation. Unlike the strain

complemented with the native bcsA_401, such a mutant could not

restore the self‐flocculating phenotype of ZM401. These results

indicated the importance of the PilZ domain for cellulose synthase to

be functional properly in cellulose biosynthesis. As a result, we deduce

that the intracellular level of c‐di‐GMP would influence the expression

of cellulose synthase and its role in cellulose biosynthesis for the self‐flocculation of Z. mobilis, which should be investigated in the future.

Based on the role of bcsA_401 in the self‐flocculation of ZM401, we

engineered ZM4 with the overexpression of this gene, but unfortu-

nately engineered strains showed very weak self‐flocculation with their

flocculating efficiencies about 35% only. Take into consideration of the

bcs operon usually function together, we further overexpressed the

whole cellulose synthase operon bcsABC from the genome of ZM401

into ZM4. Although the flocculating efficiency was increased to 72%,

the self‐flocculation was still not complete as that observed in ZM401.

Due to the difference in the genome and transcriptome profiles

between ZM4 and ZM401, other mechanism needs to be explored for

engineering ZM4 and other bacterial cells with desired self‐flocculatingphenotypes for applications in bioprocess engineering such as the self‐immobilization of bacterial cells within bioreactors for high product

F IGURE 6 Role of bcsA_401 in theself‐flocculation of ZM401. Flocculatingefficiency (a), florescence intensities

detected under the Calcofluor‐whitestaining conditions (b), and SEM images ofthe cell surfaces of ZM401ΔbcsA (c) and

the complemented mutants ofZM401ΔbcsA (d) [Color figure can beviewed at wileyonlinelibrary.com]

XIA ET AL. | 2723

productivities as well as merits in microbial physiology associated with

the morphological change, particularly the improvement in stress

tolerance for high product titers and other benefits.

4 | CONCLUSIONS

We confirmed that cellulose synthesized by ZM401 is the major EPS

contributing to its self‐flocculation. Comparative genome analysis

detected the deletion of a nucleotide thymine in ZMO1082, creating

a frame‐shift mutation and a new gene bcsA_401 through integrating

ZMO1082 into ZMO1083. We predict that the gene encodes a

protein with more transmembrane helices, and changes in the

configuration and conformation of the protein would enhance

cellulose synthesis and export, development of cellulose fibrils and

their entanglement for the self‐flocculation of ZM401.

Deactivation of both Type I and Type IV R‐M systems in Z. mobilis

was developed, which significantly improved transformation efficien-

cies of genetic modifications on the species. The de‐flocculation of

ZM401 achieved by knocking out bcsA_401 and the restoration of

the self‐flocculating phenotype to ZM401ΔbcsA through the gene

complementation experimentally supported our speculation on the

role of bcsA_401 in the self‐flocculation of ZM401. The over-

expression of bcsABC_401 in ZM4 created the self‐flocculatingphenotype further confirmed the essential role of enhanced cellulose

synthesis in the self‐flocculation of Z. mobilis. The progress will lay a

foundation for fundamental research in further deciphering mole-

cular mechanism underlying the self‐flocculation of Z. mobilis and

stress tolerance improvement associated with the morphological

change as well as for technological innovations in engineering non‐flocculating Z. mobilis and other bacterial cells with the self‐flocculating phenotype.

ACKNOWLEDGMENTS

Funding support from the National Science Foundation of China

(NSFC) with the grant reference number of (21536006) is acknowl-

edged. The assistance in the EPS analysis and characterization from

Professors Guo‐Ping Sheng and Yang Mu at the University of Science

and Technology of China, and donation of plasmids by Professors Jie

Bao at East China University of Science and Technology, Shihui Yang

at Hubei University and Ningyi Zhou at Shanghai Jiao Tong

University are highly appreciated.

CONFLICTS OF INTEREST

The authors declare that they have no conflicts of interest.

AUTHOR CONTRIBUTIONS

J. Xia designed a roadmap for this study under the supervision of

F.‐W. Bia, completed all experiments, interpreted data and developed

a manuscript draft. C.‐G. Liu supervised the culture and ethanol

fermentation of Z. mobilis. X.‐Q. Zhao supervised molecular manip-

ulations on Z. mobilis; Y. Xiao and X.‐X. Xia critically commented on

this manuscript. F.‐W. Bia supervised the whole research, analyzed

the data and other results with all co‐authors and rewrote this

manuscript.

ORCID

Feng‐Wu Bai http://orcid.org/0000-0003-1431-4839

REFERENCES

Amikam, D., & Galperin, M. Y. (2006). PilZ domain is part of the bacterial

c‐di‐GMP binding protein. Bioinformatics, 22(1), 3–6.

Asraf, S. A. K. S., Rajnish, K. N., & Gunasekaran, P. (2011). Bioinformatics

and biosynthesis analysis of cellulose synthase operon in Zymomonas

mobilis ZM4 . The IIOAB Journal, 2(3), 1–7.

Bai, F. W., Anderson, W. A., & Moo‐Young, M. (2008). Ethanol

fermentation technologies from sugar and starch feedstocks. Biotech-

nology Advances, 26, 89–105.

Benach, J., Swaminathan, S. S., Tamayo, R., Handelman, S. K., Folta‐Stogniew, E., Ramos, J. E., … Hunt, J. F. (2007). The structural basis of

cyclic diguanylate signal transduction by PilZ domains. The EMBO

Journal, 26, 5153–5166.

Chong, S., Sen, T. K., Kayaalp, A., & Ang, H. M. (2012). The performance

enhancements of upflow anaerobic sludge blanket (UASB) reactors

for domestic sludge treatment – A state‐of‐the‐art review. Water

Research, 46, 3434–3470.

Conway, T., Sewell, G. W., & Ingram, L. O. (1987). Glyceraldehyde‐3‐phosphate dehydrogenase gene from Zymomonas mobilis: Cloning,

sequencing, and identification of promoter region. Journal of Bacter-

iology, 169, 5653–5662.

Cowles, K. N., Willis, D. K., Engel, T. N., Jones, J. B., & Barak, J. D. (2015).

Diguanylate cyclases AdrA and STM1987 regulate Salmonella enterica

exopolysaccharide production during plant colonization in an envir-

onment‐dependent manner. Applied and Environmental Microbiology,

82, 1237–1248.

Dong, H. W., Bao, J., Ryu, D. D. Y., & Zhong, J. J. (2011). Design and

construction of improved new vectors for Zymomonas mobilis

recombinants. Biotechnology and Bioengineering, 108, 1616–1627.

Flemming, H. C., & Wingender, J. (2010). The biofilm matrix. Natural

Reviews Microbiology, 8, 623–633.

Flemming, H. C., Wingender, J., Szewzyk, U., Steinberg, P., Rice, S. A., &

Kjelleberg, S. (2016). Biofilms: An emergent form of bacterial life.

Natural Reviews Microbiology, 14, 563–575.

Gu, H., Zhang, J., & Bao, J. (2015). High tolerance and physiological

mechanism of Zymomonas mobilis to phenolic inhibitors in ethanol

fermentation of corncob residue. Biotechnology and Bioengineering,

112, 1770–1782.

Halme, A., Bumgarner, S., Styles, C., & Fink, G. R. (2004). Genetic and

epigenetic regulation of the FLO gene family generates cell‐surfacevariation in yeast. Cell, 116, 405–415.

He, L. Y., Zhao, X. Q., & Bai, F. W. (2012). Engineering industrial

Saccharomyces cerevisiae strain with the FLO1‐derivative gene isolated

from the flocculating yeast SPSC01 for constitutive flocculation and

fuel ethanol production. Applied Energy, 100, 33–40.

He, M., Wu, B., Qin, H., Ruan, Z., Tan, F., Wang, J., … Hu, Q. (2014).

Zymomonas mobilis: A novel platform for future biorefineries.

Biotechnology for Biofuels, 7, 101.

Hengge, R. (2009). Principles of c‐di‐GMP signaling in bacteria. Natural

Reviews Microbiology, 7, 263–273.

2724 | XIA ET AL.

Hmelo, L. R., Borlee, B. R., Almblad, H., Love, M. E., Randall, T. E., Tseng, B.

S., … Harrison, J. J. (2015). Precision‐engineering the Pseudomonas

aeruginosa genome with two‐step allelic exchange. Natural Protocols,

10, 1820–1841.

Hoang, T. T., Karkhoff-Schweizer, R. R., Kutchma, A. J., Schweizer, H. P.

(1998). A broad-host-range Flp-FRT recombination system for site-

specific excision of chromosomally-located DNA sequences: applica-

tion for isolation of unmarked Pseudomonas aeruginosa mutants.

Jeffries, T. W. (2006). Engineering yeasts for xylose metabolism. Current

Opinion in Biotechnology, 17, 320–326.

Jenal, U., Reinders, A., & Lori, C. (2017). Cyclic di‐G‐MP: Second

messenger extraordinaire. Natural Reviews Microbiology, 15, 271–284.

Jeon, Y. J., Xun, Z., Su, P., & Rogers, P. L. (2012). Genome‐wide

transcriptomic analysis of a flocculent strain of Zymomonas mobilis.

Applied Microbiology and Biotechnology, 93, 2513–2518.

Ju, F., & Zhang, T. (2015). Bacterial assembly and temporal dynamics in

activated sludge of a full‐scale municipal wastewater treatment plant.

The ISME Journal, 9, 683–695.

Kalnenieks, U. (2006). Physiology of Zymomonas mobilis: Some unan-

swered questions. Advances in Microbial Physiology, 51, 73–117.

Kerr, A. L., Jeon, Y. J., Svenson, C. J., Rogers, P. L., & Neilan, B. A. (2011).

DNA restriction‐modification systems in the ethanologen Zymomo-

nas mobilis ZM4. Applied Microbiology and Biotechnology,

89, 761–769.

Kubiak, K., Kurzawa, M., Jędrzejczak‐Krzepkowska, M., Ludwicka, K.,

Krawczyk, M., Migdalski, A., … Bielecki, S. (2014). Complete genome

sequence of Gluconacetobacter xylinus E25 strain‐valuable and

effective producer of bacterial nanocellulose. Journal of Biotechnology,

176, 18–19.

Li, Q., Zhao, X. Q., Chang, A. K., Zhang, Q. M., & Bai, F. W. (2012). Ethanol‐induced yeast flocculation directed by the promoter of TPS1 encoding

trehalose‐6‐phosphate synthase 1 for efficient ethanol production.

Metabolic Engineering, 14, 1–8.

Li, X., Thomason, L. C., Sawitzke, J. A., Costantino, N., & Court, D. L. (2013).

Positive and negative selection using the tetA‐sacB cassette: Recom-

bineering and P1 transduction in Escherichia coli. Nucleic Acids

Research, 41(22), e204–e204.

Lim, S. J., & Kim, T. H. (2014). Applicability and trends of anaerobic granular

sludge treatment processes. Biomass and Bioenergy, 60, 189–202.

Limoli, D. H., Jones, C. J., & Wozniak, D. J. (2015). Bacterial extracellular

polysaccharides in biofilm formation and function.Microbiol Spectrum, 3(3),

MB‐0011‐2014. https://doi.org/10.1128/microbiolspec.MB-0011-2014

Liu, X. M., Sheng, G. P., Luo, H. W., Zhang, F., Yuan, S. J., Xu, J., … Yu, H. Q.

(2010). Contribution of extracellular polymeric substances (EPS) to

the sludge aggregation. Environmental Science and Technology, 44,

4355–4360.

Mazur, O., & Zimmer, J. (2011). Apo‐ and cellopentaose‐bound structures

of the bacterial cellulose synthase subunit BcsZ*. Journal of Biological

Chemistry, 286, 17601–17606.

Mcnamara, J. T., Morgan, J. L. W., & Zimmer, J. (2015). A molecular

description of cellulose biosynthesis. Annual Review of Biochemistry,

84, 895–921.

Morgan, J. L. W., McNamar, J. T., & Zimmer, J. (2015). Mechanism of

activation of bacterial cellulose synthase by cyclic di‐GMP. Nature

Structural & Molecular Biology, 21, 489–496.

Morgan, J. L. W., Strumillo, J., & Zimmer, J. (2013). Crystallographic

snapshot of cellulose synthesis and membrane translocation. Nature,

493, 181–186.

Palha, M. A. P. F., Lopes, C. E., & Pereira, N., jr. (1997). Ethanol stimulates the

flocculation of Zymomonas mobilis. Biotechnology Letters, 19, 499–501.

Papenfort, K., & Bassler, B. L. (2016). Quorum sensing signal‐responsesystems in Gram‐negative bacteria. Natural Reviews Microbiology, 14,

576–588.

Seo, J. S., Chong, H., Park, H. S., Yoon, K. O., Jung, C., Kim, J. J., … Kang, H.

S. (2005). The genome sequence of the ethanologenic bacterium

Zymomonas mobilis ZM4. Nature Biotechnology, 23, 63–68.

Serra, D. O., Richter, A. M., & Hengge, R. (2013). Cellulose as an

architectural element in spatially structured Escherichia coli biofilms.

Journal of Bacteriology, 195, 5540–5554.

Soares, E. V. (2010). Flocculation in Saccharomyces cerevisiae: A review.

Journal of Applied Microbiology, 110, 1–18.

Verstrepen, K. J., Jansen, A., Lewitter, F., & Fink, G. R. (2005). Intragenic

tandem repeats generate functional variability. Nature Genetics, 37,

986–990.

Xue, C., Zhao, X. Q., & Bai, F. W. (2010). Effect of the size of yeast flocs

and zinc supplementation on continuous ethanol fermentation

performance and metabolic flux distributions under very high

concentration conditions. Biotechnology and Bioengineering, 105,

935–944.

Yang, S., Mohagheghi, A., Franden, M. A., Chou, Y. C., Chen, X., Dowe, N., …

Zhang, M. (2016). Metabolic engineering of Zymomonas mobilis for

2,3‐butanediol production from lignocellulosic biomass sugars. Bio-

technology for Biofuels, 9, 189.

Yang, S., Pappas, K. M., Hauser, L. J., Land, M. L., Chen, G. L., Hurst, G. B., …

Brown, S. D. (2009). Improved genome annotation for Zymomonas

mobilis. Nature Biotechnology, 27(10), 893–894.

Yang, S., Fei, Q., Zhang, Y. P., Contreras, L. M., Utturkar, S. M., Brawn, S. D., &

Zhang, M. (2016). Zymomonas mobilis as a model system for production

of biofuels and biochemical. Microbial Biotechnology, 9, 699–717.

Zhang, K., Shao, H., Cao, Q., He, M., Wu, B., & Feng, H. (2015).

Transcriptional analysis of adaptation to high glucose concentrations

in Zymomonas mobilis. Applied Microbiology and Biotechnology, 99,

2009–2022.

Zhang, M., Eddy, C., Deanda, K., Finkelstein, M., & Picataggio, S. (1995).

Metabolic engineering of a pentose metabolism pathway in ethanolo-

genic Zymomonas mobilis. Science, 267, 240–243.

Zhao, N., Bai, Y., Liu, C. G., Zhao, X. Q., Xu, J. F., & Bai, F. W. (2014).

Flocculating Zymomonas mobilis is a promising host to be engineered

for fuel ethanol production from lignocellulosic biomass. Biotechnology

Journal, 9, 362–371.

Zhao, N., Bai, Y., Zhao, X. Q., Yang, Z. Y., & Bai, F. W. (2012).

Draft genome sequence of the flocculating Zymomonas mobilis

strain ZM401 (ATCC 31822). Journal of Bacteriology, 194,

7008–7009.

Zhao, X. Q., & Bai, F. W. (2009). Yeast flocculation: New story in fuel

ethanol production. Biotechnology Advances, 27, 849–856.

How to cite this article: Xia J, Liu C-G, Zhao X-Q, Xiao Y, Xia

X-X, Bai F-W. Contribution of cellulose synthesis, formation of

fibrils and their entanglement to the self‐flocculation of

Zymomonas mobilis. Biotechnology and Bioengineering.

2018;115:2714–2725. https://doi.org/10.1002/bit.26806

XIA ET AL. | 2725